I did not realize that functions can fill in for predicates directly without lambda notation. To illustrate, consider the following:

void Main()
{
	var words = new List<string>()
	{
		"therapists",
		"s words",
		"slang",
		"mustache",
		"sean connery"
	};

	var s_words = words.Where(w => w.StartsWith("s" ,StringComparison.CurrentCultureIgnoreCase));

	foreach(var word in s_words)
	{
		Console.WriteLine(word);
	}
}

It works, but it’s a rather long lambda, which is why I broke it out of the for loop. Let’s put that logic into a function, getting this:

void Main()
{
	var words = new List<string>()
	{
		"therapists",
		"s words",
		"slang",
		"mustache",
		"sean connery"
	};

	foreach(var word in words.Where(w => SWords(w)))
	{
		Console.WriteLine(word);
	}
}

public bool SWords(string word)
{
	return word.StartsWith("s" ,StringComparison.CurrentCultureIgnoreCase);
}

That’s better. And you might find yourself using your test in other places in the code, so it’s useful to have the function. What I found out recently, is that you can go one step further:

void Main()
{
	var words = new List<string>()
	{
		"therapists",
		"s words",
		"slang",
		"mustache",
		"sean connery"
	};

	foreach(var word in words.Where(SWords))
	{
		Console.WriteLine(word);
	}
}

public bool SWords(string word)
{
	return word.StartsWith("s" ,StringComparison.CurrentCultureIgnoreCase);
}

In this example, the savings may not look drastic. But for several chained methods you can gain a lot of brevity and clarity.

Post to Twitter